The Earl of Rosse un the Radiation of Heat from the Moon. 315 



(2) That which falls from the sun on the moon's surface, and 

 is at once reflected regularly and irregularly. 



(3) That which, falling from the sun on the moon's surface, is ab- 

 sorbed, raises the temperature of the moon's surface, and is after- 

 wards radiated as heat of low refrangibility. 



The apparatus consisted of a thermopile of four elements, the faces 

 half an inch square, on which all the moon's heat which falls on 

 the large speculum of the 3-foot telescope is concentrated, by means 

 of a concave mirror of 3^- inches aperture, 2*8 inches focal length. 



As it was found difficult to compensate the effects of unequal ra- 

 diation on the anterior face of the pile, by exposing the posterior 

 face also of the same pile to radiation from the sky, during the 

 later experiments (beginning with March 23rd) two piles were used, 

 and the following was the form of apparatus adopted. 



D E is the large mirror of the telescope ; F G the two small 

 concave mirrors of 3| inches aperture, and 2*8 inches focal length, 

 fixed in the plane of the image formed by the large mirror D E. 

 The two thermopiles are placed respectively in the foci of F and 

 G, their anterior faces shielded from wind and other disturbing 

 causes by polished brass cones, and their posterior faces kept at a 

 nearly uniform temperature by means of brass caps filled with 

 water. The thermopiles and accompanying mirrors are supported 

 by a bar screwed temporarily on the mouth of the tube. Two wires 

 are connected with the two poles of each pile ; and the ends of the 

 wires are connected,, two and two, close to the galvanometer, in such 

 a manner that a given amount of heat on the anterior face of one 

 pile will produce a deviation equal in amount, and opposite in direc- 

 tion, to that produced by an equal amount of heat on the anterior 

 face of the other pile. Thomson's Reflecting Galvanometer was the 

 one used. 



This apparatus has not yet had a fair trial, as I was unable to 

 obtain from Messrs. Elliot a pile ready made of similar dimensions 

 to that which I already possessed. That which they sent had only 

 one-fourth the required area of face. 



The following Table (p. 316) is a summary of the results. 



In column 3 is given the mean of the deviations of all the single 

 differences from the mean difference of all the readings taken with 

 the moon on and with the moon off the apparatus. 



In column 4 the arithmetic mean of all the observed deviations. 



In column 5 the calculated deviation for each night at midnight, 

 on the assumption that the deviation corresponding to full moon 

 = 100, and that the moon is a smooth sphere. 



Y2 



